From d8ef33a551bc6189cfd93777571eb8db3cf79c94 Mon Sep 17 00:00:00 2001 From: jianzhao-xu <978716854@qq.com> Date: Tue, 24 Feb 2026 12:25:20 +0800 Subject: [PATCH] bugfix: prioritize init_npu_backend to fix various initialization bugs (#17652) --- python/sglang/srt/model_executor/model_runner.py | 5 ----- python/sglang/srt/server_args.py | 7 ++++++- 2 files changed, 6 insertions(+), 6 deletions(-) diff --git a/python/sglang/srt/model_executor/model_runner.py b/python/sglang/srt/model_executor/model_runner.py index 16a28f099..cdb877416 100644 --- a/python/sglang/srt/model_executor/model_runner.py +++ b/python/sglang/srt/model_executor/model_runner.py @@ -193,11 +193,6 @@ _is_npu = is_npu() _is_cpu_amx_available = cpu_has_amx_support() _is_cpu_arm64 = is_host_cpu_arm64() -if _is_npu: - from sglang.srt.hardware_backend.npu.utils import init_npu_backend - - init_npu_backend() - MLA_ATTENTION_BACKENDS = [ "aiter", "flashinfer", diff --git a/python/sglang/srt/server_args.py b/python/sglang/srt/server_args.py index 83c7d425e..a1bbd2038 100644 --- a/python/sglang/srt/server_args.py +++ b/python/sglang/srt/server_args.py @@ -885,7 +885,12 @@ class ServerArgs: def _handle_npu_backends(self): if self.device == "npu": - from sglang.srt.hardware_backend.npu.utils import set_default_server_args + from sglang.srt.hardware_backend.npu.utils import ( + init_npu_backend, + set_default_server_args, + ) + + init_npu_backend() set_default_server_args(self)